Keep Claude Code context clean. Open-source toolkit: drift detection, re-read dedup, integrity scoring, AST-aware reads, 15 MCP tools. 62.6% measured savings, reproducible.
Skill Claude CodeCodex
Run the published 20-task Sipcode benchmark on the user's machine and report measured median savings (target 62.6%). Reproducible, locked corpus. Use when the user asks "what's the proof?", "how much does sipcode actually save?", "run the benchmark", or wants to verify the headline savings claim themselves.
Skill Claude CodeCodex
Predict the dollar cost of a coding task across Opus 4.7, Sonnet 4.6, and Haiku 4.5 BEFORE the user runs it. Pure heuristic + historical-anchor math, zero LLM calls. Use when the user asks "how much will X cost?", "is this task worth doing on Opus?", "estimate the refactor", or any pre-flight pricing question.
Skill Claude CodeCodex
A/B compare the user's actual token spend before and after they installed Sipcode's optimizers — on their own session data. The honesty-first savings verifier. Use when the user asks "is sipcode actually saving me tokens?", "show me the impact", "prove sipcode is working", or any variant of "is this thing worth…
Skill Claude CodeCodex
Forensic audit of the user's most recent Claude Code session. Returns total spend, output ratio, duplicate file reads, idle context after auto-compact, top expensive tool calls, and an estimate of what Sipcode would have RECOVERED if optimizers were active. Use when the user asks "where did my tokens go?", "audit my…